Closed loop system

I do not use a closed loop. However, your experience does not put me off because it is known that CGMs are inaccurate at high BG and closed loop should aim to avoid high BG.

It is important not to rely on CGM readings when high check with a finger prick to avoid hypos when correcting from inaccurate readings.

I am (sort of) using a closed loop system but as it doesn't take account of raised insulin insulin resistance once your bg is over 10mmol/L I have to manually correct when I go over that anyway. In fact, it has so many safeguards built in to avoid overdosing that I'm pretty sure I wouldn't get a serious overdose from a false reading unless I manually put it in. My dexcom can read low for the first 24 hours so I put my pump in manual mode when this happens.... (Just had a sensor fail on start up so was unable to pre insert it so went manual for a day.)

But I am always skeptical about very high or very low bg readings from a sensor (learnt from bitter experience when on pens and my dexcom said I was low and I kept knocking back the carbs and went up to 18.) my pump and phone both generate large numbers of alerts once I go seriously out of range so I'd have to test manually then anyway.

I'm on the tandem tslim x2. I did worry slightly about this issue before going onto a pump but I think that (this pump at least) just isn't as closed loop as you might believe. Strange readings require manual oversight.

My viewpoint (been on HCL for 1.5 years and now FCL for .5 year)

All the tech is complex - there are 4 points of failure - Pump, Pump controller (PDM, Phone, software on pump etc), CGM and Cannula. Any one of those can go pear shaped - most common is the CGM

I've had a couple of CGMs do odd things like very high or low readings before failing or flat lining.

The only way to get past this is to check when things don't seem right - you begin to know your patterns for how the loop works, and if it seems to have changed radically - check the cannula and CGM are actually working as expected (I have experienced resistance from a not-happy cannula inserted.

My Loop will pick up consistently bad CGM readings and flag it as in error - but it's not perfect.

Benefits are looping is so much better than non-looping for me - takes the chimpanzee hitting buttons (me) completely out the loop, as long as the chimp keeps an eye on stuff going bananas.

We will always get problems with the tech - not often, but they will happen. For me its many time less problems than me getting things wrong and over bolusing, or forgetting to do stuff to cover unforeseen events (like a surprise cream cake) - Worth the risk of the occasion Libre 2 going nuts and giving me a headache. If the L2 gives low readings - my loop suspends insulin delivery and alarms - as safe as I think is possible for the freedom it gives - we all pick our own path though

sebg said:

My libre 2 CGM has been given some false readings recently for example my scan showed I was on 17.3 so I injected based on this having used a CGM for a couple of years now I didn't question it, this caused a hipo and since then I have been doing finger prick test more often to make sure that I don't overdose again.

This has changed my enthusiasm about the closed loop system and now I'm quite frankly concerned about using one in future.

Has anyone got experience using a closed loop system and have you had any issues with false readings from the CGM

Click to expand...

I had this same concern as I get counter-regulatory (non-carb) spikes in the night that rise to 12-14 and these are insulin resistant. My high-glucose alarm wakes me and I can inject 10-15 units fast-acting before I see any effect and sometimes the insulin sensitivity returns and my BG plummets. This is analagous to your CGM false reading that could result in too much insulin.

If all of this were on a closed-loop system then a low BG could occur but you would set low-glucose alarms to wake you up, just like normal.

Bottom line is a closed-loop system doesn't remove you from the picture - it is not a plug-in walk-away device, and we would still need to monitor BG and adjust or treat highs and lows. But for most of the time your control will be significantly better.

I’m on Omnipod 5 and Dexcom G6. When ever I put on a new sensor, I always put Omnipod into manual mode and wait for the CGM readings to become stable before going back into automatic made. If I have rouge readings, before doing anything I confirm with a finger prick test. I’ve never had a new sensor come on line and not need a few calibrations before Settling down 24 hrs later. Will be interesting to see as will be swapping to the Libre2 + sensor at some stage, although I’m not looking forward to not having the ability to calibrate

RD02 said:

Will be interesting to see as will be swapping to the Libre2 + sensor at some stage, although I’m not looking forward to not having the ability to calibrate

Click to expand...

I calibrate my L2 it using xDrip. Currently xDrip doesn't support the L2+ (its in development) but there are work arounds. If you find it essential - there are hacks to get it done

sebg said:

I also get spikes in the night usually 2 hours before I want to get up and I would give myself 10-15 units without food and go back to sleep but since I've been getting false readings (with my last two sensors) I have just been going back to sleep without injecting as it's only a couple of hours until I get up anyway
I've lost trust in the readings and have been doing finger prick test any time I injected or get an alarm I just don't trust the machine to work out how much and when I need insulin

Click to expand...

It's not a good sign when you lose trust in the devices and have to worry about health consequences.
I think that's the Dawn Phenomenon you're taking about, and mine starts at 3:00-3:30 am. Like you I give myself insulin and set my alarm for 4.5 mmol/L (and set my clock alarm for 5 am because I'm also not convinced the CGM alarm is working properly).

Will diabetics ever get an uninterrupted night's sleep?